Nuclear receptor liver receptor homologue 1(LRH-1) regulates pancreatic cancer cellgrowth and proliferationCindy Benoda, Maia V. Vinogradovaa, Natalia Jouravela, Grace E. Kimb, Robert J. Flettericka,1, and Elena P. Sablina

aDepartment of Biochemistry and Biophysics, University of California 600 16th Street, GH S412E, San Francisco, CA 94158-2517; andbDepartment of Pathology, University of California 505 Parnassus Avenue, M561, San Francisco, CA 94143-0102

Contributed by Robert J. Fletterick, August 18, 2011 (sent for review May 23, 2011)

An essential regulator of gene transcription, nuclear receptor liverreceptor homologue 1 (LRH-1) controls cell differentiation in thedeveloping pancreas and maintains cholesterol homeostasis inadults. Recent genome-wide association studies linked mutationsin the LRH-1 gene and its up-stream regulatory regions to develop-ment of pancreatic cancer. In this work, we show that LRH-1 tran-scription is activated up to 30-fold in human pancreatic cancer cellscompared to normal pancreatic ductal epithelium. This activationcorrelates with markedly increased LRH-1 protein expression inhuman pancreatic ductal adenocarcinomas in vivo. Selective block-ing of LRH-1 by receptor specific siRNA significantly inhibits pan-creatic cancer cell proliferation in vitro. The inhibition is trackedin part to the attenuation of the receptor’s transcriptional targetscontrolling cell growth, proliferation, and differentiation. Pre-viously, LRH-1 was shown to contribute to formation of intestinaltumors. This study demonstrates the critical involvement of LRH-1in development and progression of pancreatic cancer, suggestingthe LRH-1 receptor as a plausible therapeutic target for treatmentof pancreatic ductal adenocarcinomas.

protein target ∣ gene regulation

With mortality rate nearing its incidence, pancreatic ductaladenocarcinoma (PDAC) presents a challenge for modern

oncology. Current chemotherapy drugs approved for pancreaticcancer are not organ specific and are modestly effective. Thus,there is a need for improved therapeutic options and effectivepancreatic cancer drugs. Recent studies reveal that signalingpathways are similar in pancreatic development and malignantgrowth in the adult pancreas (1, 2). One of the common drivingfactors in pancreatic embryo- and oncogenesis is the nuclearreceptor liver receptor homologue 1 (LRH-1, NR5A2; ref. 3). Inadults, this protein is expressed primarily in liver, intestine, andpancreas, where it regulates expression of proteins maintainingcholesterol homeostasis (3, 4). LRH-1 is also expressed in theovary and breast adipose tissue where it controls biosynthesis ofsteroids (5, 6). LRH-1 is vital in early development; it maintainsundifferentiated ES cells by controlling expression of two tran-scription factors, Oct4 and Nanog (7–9). Recent studies demon-strated that LRH-1 can substitute for Oct4 in derivation ofinduced pluripotent stem (iPS) cells (9). LRH-1 is classed as anorphan nuclear receptor because its activating hormones havenot been identified. Crystallographic and biochemical studiespresented compelling evidence that LRH-1 could bind ligands(10–13) and suggested phosphatidyl inositols as potential hor-mone candidates for this receptor (10). LRH-1 is also regulatedvia posttranslational modifications, phosphorylation, and sumoy-lation (14, 15). Known transcriptional regulators of LRH-1include steroid receptor coactivators (SRCs), the CREB-bindingprotein (CBP), and the peroxisome proliferator-activated recep-tor gamma coactivator 1-alpha (PGC-1α) as well as the corepres-sors silencing mediator for retinoic acid and thyroid hormonereceptor (SMRT), short heterodimer partner (SHP), the pros-pero homeobox protein 1 (Prox1), and the orphan nuclear recep-

tor dosage sensitive sex reversal, adrenal hypoplasia criticalregion on chromosome X gene 1 (Dax-1) (3, 16, 17). No syntheticantagonists of LRH-1 are available to date. LRH-1 is linked tomultiple developmental pathways including Wnt/β-catenin (8,18), Hedgehog signaling (19), and regulatory cascades controlledby the pancreas-specific transcription factor pancreatic and duo-denal homeobox 1 (PDX-1) (20). In pancreatic cancer, thesepathways are reactivated and shown to be essential for develop-ment of pancreatic tumors (1, 2, 21, 22). Recent genome-wideassociation studies link mutations in the LRH-1 gene and itsup-stream regulatory regions to development of PDAC (23).

In this study, we demonstrate that transcription of LRH-1gene is misregulated, and its expression is activated in humanPDACs compared to normal adult pancreas. We show that theaberrant expression of LRH-1 in pancreatic cancer cells can beattenuated by receptor specific siRNA. This suppression resultsin significant inhibition of cancer cell proliferation, suggestingthat LRH-1 has a major role in development of pancreatic can-cer. This finding complements the previous report implicatingLRH-1 in formation of intestinal tumors (24).

ResultsIdentification of LRH-1 in Pancreatic Cancer Cells Cultured in Vitro.Using quantitative PCR (qPCR) and Western blot analyses, weshowed that LRH-1 receptor is present in nine well-characterizedpancreatic cancer cell lines derived either from the primaryPDAC tumors or metastatic sites (Fig. 1). The receptor levels inmost tested primary tumor cells were comparable to those foundin nonneoplastic pancreatic ductal epithelium (Hpde6c7, Fig. 1,Left). In contrast, in most metastatic pancreatic cancer cells,the levels of the receptor transcripts were significantly increased(25–30-fold change in AsPC-1, COLO 357, and its derivative cellline L3.6 pl; Fig. 1, Right). In contrast to COLO 357 and L3.6 pl,in cells L3.3 (another derivative of COLO 357) the LRH-1 tran-scription appeared to be disabled. Western blot analysis confirmedthe presence of LRH-1 in primary cancer cells and revealed itssignificantly higher expression in cells originated from metastatictumors (shown at the bottom of Fig. 1). No LRH-1 protein wasdetected in L3.3 cells; because of the apparent absence of the re-ceptor in these cells, we used them as a negative control in ourfurther experiments. In four cancer cell lines, the activated tran-scription of LRH-1 gene correlated with reappearance of PDX-1, a known transcriptional regulator of the receptor (20). However,no clear reciprocity was revealed between these two proteins(Fig. S1 A and B). Notably, we observed significantly decreased

Author contributions: R.J.F. and E.P.S. designed research; C.B., M.V.V., and N.J. performedresearch; C.B. and M.V.V. contributed new reagents/analytic tools; C.B., M.V.V., N.J., G.E.K.,R.J.F., and E.P.S. analyzed data; G.E.K. interpreted immunohistochemistry pathology data;and C.B., M.V.V., R.J.F., and E.P.S. wrote the paper.

The authors declare no conflict of interest.1To whom correspondence should be addressed. E-mail: Robert.Fletterick@ucsf.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1112047108/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1112047108 PNAS ∣ October 11, 2011 ∣ vol. 108 ∣ no. 41 ∣ 16927–16931

APP

LIED

BIOLO

GICAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

May

1, 2

021

levels of Prox1, an established corepressor of LRH-1 (16), in eightout of 10 cancer cell lines (Fig. S1C). We also found the transcriptsfor SHP, another corepressor of LRH-1 (3, 11), diminished inseven cancer cell lines tested (Fig. S1D).

Expression of LRH-1 in Human PDAC Tumors.We analyzed the expres-sion of the receptor in human resection specimens obtained fromnormal and cancerous pancreas. Twelve independent cases ofPDAC and four samples of normal pancreas were studied usingLRH-1 specific immunohistochemistry (IHC) staining with tworeceptor-specific antibodies [HPA005455 by Sigma-Aldrich andLS-A2447 by LifeSpan Biosciences; the specificity of antibodieswas confirmed using positive (liver) and negative (heart muscletissue] controls (Fig. S2). The IHC experiments revealed cleardifferences between the levels of LRH-1 in normal and cancerouspancreas (Fig. 2 and Fig. S2; for statistical analysis of IHC data,see Table S1). In normal exocrine pancreas, the LRH-1 proteinwas present at relatively low levels. Both acini and ducts showedweak LRH-1 staining in the cytoplasm and weak-to-moderatepunctate staining in the nuclei. In contrast, in all tested PDACs,the presence of LRH-1 was markedly increased, with heightenedlevels of the protein detected either in the nuclei or the cytoplasmof cancerous cells. In some neoplastic cells, the receptor ap-peared to localize predominantly at the cytoplasm. Whereas thefibroblasts in the stroma surrounding the neoplastic lesions didnot show significant staining, many infiltrating immune cells werepositively stained for LRH-1. An increased presence of LRH-1was also detected in the acinar cells affected by pancreatitis aswell as in pancreatic intraepithelial neoplasia (PanIN) lesions(Fig. 2 and Fig. S2, indicated).

Blocking LRH-1 by Specific siRNA. We analyzed whether the LRH-1transcripts and protein levels in pancreatic cancer cells are dimin-ished after treatments with receptor specific siRNA. Two differ-ent siRNAs corresponding to the receptor N-terminal region andits ligand-binding domain [nucleotides 411–431 (siRNA-1) and1504–1524 (siRNA-2); ref. 25] were used for these experiments.Quantitative PCR analysis showed nearly complete inhibition ofLRH-1 gene transcription on the second day after the treatments(Fig. 3A; the transcript levels returned to the original by day 4likely due to siRNA degradation). In cells treated with controlsiRNA, the transcription of LRH-1 was not affected. Comple-menting qPCR data, Western blot analysis showed that expres-sion of LRH-1 protein in cells treated with specific siRNAs wassubstantially reduced; treatment with control siRNA had noeffect on LRH-1 expression (Fig. 3B).

Inhibition of Pancreatic Cancer Cell Proliferation by Anti-LRH-1 siRNA.Using siRNA as an efficient tool for silencing the receptor,we analyzed the role of LRH-1 in proliferation of pancreatic can-cer cells. These experiments revealed that specific blocking ofLRH-1 results in significant inhibition of cancer cell prolifera-tion; inhibition was observed in four different cell lines expressingthe receptor (Fig. 3C, Left). Notably, proliferation of the samecells treated with control siRNA was not affected. Furthermore,proliferation of pancreatic cancer cells L3.3, which do not expressLRH-1 (Fig. 1), was not affected either (Fig. 3C, Right).

Transcription of LRH-1 Regulated Gene Targets. We tested possiblemechanisms of LRH-1 involvement in pancreatic cancer by eval-uating the levels of mRNA for its established targets, includingcyclins D1, E1 (Cyc D1, Cyc E1), and C-Myc genes (9, 18), whichare essential for cell growth and proliferation (26–29). Quantita-tive PCR data revealed that C-Myc gene is up-regulated in alltested pancreatic cancer cells (Fig. 4A). Quantitative PCR ana-lyses also showed that transcription of C-Myc was compromisedin pancreatic cancer cells treated with LRH-1 specific siRNAs(Fig. 4B). Specific inhibition of LRH-1 also resulted in dimin-ished transcription of cyclins D1 and E1 (Fig. 4B). Treatment ofcells with control siRNA had no effect on transcription of any ofthe tested targets. All measurements were performed between 24and 48 h after the transfections, when the levels of LRH-1 wereminimal (Fig. 3). Complementing these analyses, we evaluated

Fig. 1. Expression of LRH-1 in pancreatic cancer cells. Levels of LRH-1 mRNAin pancreatic cancer cells originated from primary and metastatic PDACs(Left, Right, in light and dark gray) are shown relative to control (nonneo-plastic pancreatic ductal epithelial cells Hpde6c7, in white); note 10-fold scalechange in the right panel. (Left) For comparison, the levels of LRH-1 mRNA inhepatocarcinoma cells HepG2 are shown in white. Standard deviations aredrawn as black lines. Western blots at the bottom of the panels show normal-ized expression of LRH-1 protein in the corresponding cells (imaged byenhanced chemiluminescence, 1-min exposure). The expression levels ofLRH-1 are normalized to those of β-actin (detected after 10-s exposure).

Fig. 2. Differential expression of LRH-1 in normal and cancerous pancreas.Shown are IHC images of human normal (Normal) and neoplastic pancreas(PDAC 1–3). The samples were treated with primary antibody HPA005455(Sigma-Aldrich), stained with ImmPACT diaminobenzidine (Vector Labs)and counterstained with hematoxylin; the original magnification (×40 or×10) is specified. Weak cytoplasmic and weak-to-moderate punctate nuclearstaining is observed for all components of normal exocrine pancreas. Inneoplastic cells, elevated levels of LRH-1 are observed either in the nuclei(PDAC 1) or in the cytoplasm (PDAC 2) or both. Whereas the fibroblasts sur-rounding neoplastic lesions do not show significant LRH-1 specific staining(PDAC 3), many infiltrating immune cells (indicated) are positively stainedfor LRH-1. Heightened levels of LRH-1 are also observed in PanIN lesions(indicated) and in the acinar and ductal cells affected by pancreatitis (indi-cated by arrow).

16928 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1112047108 Benod et al.

Dow

nloa

ded

by g

uest

on

May

1, 2

021

the levels of other known gene targets of LRH-1 that control celldifferentiation (Oct4 and Nanog; refs. 7–9). Although the levelsof Oct4 in pancreatic cancer cells were comparable to thatobserved in nonneoplastic ductal epithelium (Fig. S3A), Nanoggene was found to be activated in five cancer cell lines tested(Fig. S3B).

Cell Cycle Arrest Following Inhibition of LRH-1 by Specific siRNA. Weanalyzed the cell cycle of AsPC-1 cells treated with either irrele-vant or LRH-1 specific siRNAs using propidium iodide (PI)DNA staining followed by fluorescence-activated cell sorting.These studies allowed assessment of normal and apoptotic nucleias well as distribution of cell populations in G0∕G1, S, and G2∕Mphases of the cell cycle. Our data revealed that G0∕G1 populationin specifically treated cells was increased, whereas the popula-tions of cells progressing through S phase and those entering

G2 phase were decreased (Fig. S4). Based on these observations,we conclude that specific blocking of LRH-1 in pancreatic cancercells induces a cell cycle arrest in G0∕G1 phase by slowing theG1∕S transition. No increase in the apoptotic cell populationwas observed for cells transfected with either control or specificsiRNAs compared to untreated AsPC-1 cells.

DiscussionIn this study, we provide evidence of the up-regulated transcrip-tion of LRH-1 gene in PDAC cells compared to nonneoplasticpancreatic ductal epithelium. Up-regulation is modest in cancercells derived from primary tumors; however, in cells originatedfrom metastatic sites, the observed levels of LRH-1 transcriptsexceed those seen in nonneoplastic epithelium 20–30 times(Fig. 1). Activation of LRH-1 does not always correlate with thereappearance of PDX-1, a known up-stream activator of LRH-1(20) (Fig. S1 A and B), suggesting that other regulatory mechan-isms, including possible genomic alterations influencing expres-sion and activity of the receptor, might be involved in theobserved up-regulation of LRH-1 in pancreatic cancer cells. Inagreement with this idea, genome-wide association study identi-fied five SNPs associated with pancreatic cancer that mappedto LRH-1 gene and its up-stream regulatory region (23). No evi-dence of alterations in the LRH-1 gene copy number was foundin independent comparative genomic hybridization (CGH) stu-dies using pancreatic cancer cells cultured in vitro (30). Our datashowed decreased levels of Prox1 and SHP, two establishedcorepressors of LRH-1 (3, 11, 16), in the majority of testedpancreatic cancer cell lines (Fig. S1 C and D), suggesting thatchanges in regulation of transcriptional activity of LRH-1 mightbe associated with pancreatic oncogenesis.

Consistent with activation of LRH-1 in pancreatic cancer cellsin vitro, we observe increased levels of LRH-1 in human PDACsamples (Fig. 2, Fig. S2, and Table S1). In accordance with thereceptor’s established function in normal exocrine pancreas (4),relatively low levels of LRH-1 are found in the nuclei and cyto-plasm of normal acini and ducts. In contrast, in PDAC tumors,the presence of LRH-1 is markedly increased, with heightenedlevels of receptor found either in the nuclei or cytoplasm of theneoplastic cells. No evidence of LRH-1 gene amplification wasfound in independent CGH studies on human PDAC tumors(31, 32) to account for the observed receptor overexpression.Interestingly, in some neoplastic lesions, the LRH-1 protein islocalized predominantly at the cytoplasm of cancerous cells(Fig. 2 and Fig. S2). Previously, similar cytoplasmic localizationwas reported for overexpressed LRH-1 receptor in colon andbreast cancer tumors (24, 25). The implications of these observa-tions are not understood. It is hypothesized that localization ofnuclear receptors in the cytoplasm of cancer cells might be linkedto nongenomic signaling essential for tumor survival and growth(33, 34). Such cytoplasmic accumulation could also be relatedto the receptor-mediated transcription and nuclear-cytoplasmicshuttling of microRNAs in neoplastic cells; these processes areoften impaired during tumorigenesis (35, 36).

Our data show that activated expression of LRH-1 is asso-ciated with cancer growth and is evident in samples of invasivepancreatic cancer (Fig. 2, Fig. S2, and Table S1). These observa-tions suggest that the increased activity of LRH-1 in cancer cellsmight correlate with their aggressive growth and exaggeratedproliferative properties. The latter hypothesis is supported bythe established role of LRH-1 in cell proliferation. The receptoris known to induce cell proliferation through the induction ofcyclins D1 and E1, an effect thought to be potentiated by inter-actions of LRH-1 with β-catenin, a key component of the Wntsignaling pathway (8, 18). In addition to cyclins D1 and E1, ex-pression of Myc genes, which are required for cell proliferationand are also regulated via Wnt/β-catenin signaling, is shown tobe controlled by LRH-1 (9, 18). Consistent with these data, we

Fig. 3. Blocking LRH-1 function by siRNA specific to the receptor. (A) Inhibi-tion of LRH-1 transcription in AsPC-1 cells. Cells were analyzed by qPCR forrelative levels of LRH-1 mRNA at days 1–4 following transfections. Control inwhite corresponds to cells treated with irrelevant siRNA (see Materials andMethods); light- and dark-gray bars indicate mRNA in cells transfected withtwo LRH-1 specific siRNAs. (B) Western blot analyses of AsPC-1 and COLO 357cells transfected with anti-LRH-1 siRNAs. Lanes: 1 and 2, cells treated withtwo LRH-1 specific siRNA; 3, cells transfected with control siRNA; 4, nontrans-fected cells. Data are shown for day 2 following the transfections. Expressionof β-actin was used as an internal control. (C) Effect of LRH-1 knockdown oncancer cell proliferation. Proliferation of four pancreatic cancer cells expres-sing LRH-1 (indicated in the left panel) was compared following their trans-fections with two LRH-1 specific siRNAs (in light and dark gray). Data areshown for day 4 relative to the control (white). Proliferation of cells L3.3 thatdo not express LRH-1 was not affected by the anti-LRH-1 siRNAs (shown in theright panel relative to the control). The efficiency of transfections (>90%)was optimized by monitoring proliferation of cells treated with AllStarsHs Cell Death Control siRNA (in black).

Fig. 4. Effects of blocking LRH-1 on transcription of genes controlling cellproliferation. (A) Quantitative PCR data showing levels of C-Myc mRNA.Levels of C-Myc in primary and metastatic cells are shown as light- anddark-gray bars, relative to that found in nonneoplastic cells Hpde6c7. Stan-dard deviations are drawn as black lines. (B) Inhibition of C-Myc, Cyc D1, andCyc E1 gene transcription in AsPc-1 cells treated with anti-LRH-1 siRNA. Cellswere analyzed by qPCR for the relative levels of mRNA corresponding toC-Myc, Cyc D1, and Cyc E1 (indicated). Controls in white correspond to cellstreated with irrelevant siRNA; light- and dark-gray bars indicate the levels ofcorresponding mRNA in cells transfected with two LRH-1 specific siRNAs.

Benod et al. PNAS ∣ October 11, 2011 ∣ vol. 108 ∣ no. 41 ∣ 16929

APP

LIED

BIOLO

GICAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

May

1, 2

021

observe decreased levels of C-Myc as well as Cyc D1 and Cyc E1as a result of specific blocking of LRH-1 (Fig. 4). Importantly, thediminished levels of LRH-1 and its transcriptional targets corre-late with significant inhibition of pancreatic cancer cell prolifera-tion (Fig. 3C). Analysis of underlying mechanisms of LRH-1mediated control over cell proliferation revealed that blockingthe receptor affects the cell cycle by slowing the G1∕S transition(Fig. S4). This finding is congruent with the described role ofLRH-1 as a regulator of cyclins D1 and E1, which induce progres-sion through the cell cycle and stimulate the G1∕S transition (18).

Because LRH-1 is critical for maintenance of pluripotent stemcells (7, 8) and is shown to mediate reprogramming of somaticcells (9), we examined levels of most common markers associatedwith dedifferentiation and pluripotency, including transcriptionfactors Oct4 and Nanog, the established targets of LRH-1 (7–9).Although we did not find significant differences in Oct4 transcrip-tion between normal and neoplastic cells (Fig. S3A), we founddramatic amplification of Nanog transcripts in a subset of cancercells (Fig. S3B); the reappearance of Nanog in these cells might beindicative of their dedifferentiated state. The receptor-mediatedactivation of genes controlling cell growth, proliferation, and dif-ferentiation (Fig. 5) suggest that aberrant expression of LRH-1might be linked to proliferation of dedifferentiated pancreaticcancer cells, which are associated with aggressive pancreatictumors (37). Recent studies demonstrated the inherent plasticityof pancreatic cells that can be reprogrammed to the transient de-differentiated progenitor-like states, which can give rise to PDAC(38–40). Because of the established roles of LRH-1 in pancreaticand stem cell differentiation (7–9, 20), we hypothesize that thisreceptor might contribute to the reprogramming events in theadult pancreas that lead to its transformation. This hypothesisimplies that LRH-1 might be involved in maintenance of pancrea-tic cancer stem cells (37) that are thought to be the major contri-butor to the relapse of pancreatic cancer and its metastasis.

While investigating the diverse and intersecting mechanismsof the receptor involvement in pancreatic oncogenesis, we noted

that tumor-associated infiltrating immune cells, including lym-phocytes and neutrophils, also express high levels of LRH-1(Fig. 2 and Fig. S2). Similar expression of LRH-1 was previouslydetected in lymphocytes infiltrating breast carcinomas (25).These observations are captivating as accumulating data indicatethat infiltration of solid tumors by immune cells might enhancethe metastatic potential of cancer cells by facilitating their inva-sion into surrounding tissues (41). Notably, human neutrophilsproduce high levels of interleukin 1 receptor antagonist (IL-1RA), an established transcriptional target of LRH-1 (42, 43).This protein modulates the inflammatory responses during carci-nogenesis, and its overexpression correlates with poor prognosisin patients with diverse malignancies (44). LRH-1 also regulatesthe extraadrenal biosynthesis of glucocorticoids, which modulatelocal inflammatory responses (45) and prolong the survival ofneutrophils by delaying their apoptosis (46). These findings pointto another possible role for LRH-1 in pancreatic tumorigenesis,through its impact on inflammatory pathways. Previously, LRH-1was shown to contribute to intestinal tumor formation througheffects on cell cycle and inflammation (24). In agreement withthese data, we observe heightened expression of LRH-1 in pan-creatic tissue affected by pancreatitis and in PanIN lesions (Fig. 2and Fig. S2).

In summary, our work reveals that nuclear receptor LRH-1 isup-regulated and overexpressed in neoplastic pancreas comparedto normal pancreatic tissues. We show that specific blocking ofLRH-1 by siRNA inhibits pancreatic cancer cell proliferationin vitro; the inhibition is tracked to the attenuation of the recep-tor’s transcriptional targets controlling cell growth, proliferation,and differentiation. Our data, combined with recent results ofgenome-wide association study (23), demonstrate that LRH-1receptor has a major role in development of pancreatic cancer.Our study also suggests that LRH-1 might be a plausible thera-peutic target for treatment of PDAC. Although many otherregulatory molecules have been identified as the importantplayers in pancreatic tumorigenesis, critical factors contributingto the LRH-1 appeal as a drug target include the receptor’s well-defined ligand-binding pocket, which could be targeted by speci-fic synthetic modulators, and the proven efficiency of nuclear re-ceptor inhibitors for treatment of different types of malignancies,including breast and prostate cancer.

Materials and MethodsCells and Culture Conditions. Pancreatic cancer cells and immortalizedpancreatic ductal epithelium cells were kindly provided by M. McMahon[University of California, San Francisco (UCSF)]. Cells were cultured at 37 °Cin a humidified atmosphere containing 5% CO2, in DMEM supplementedwith 10% FBS, 1% L-glutamine, and antibiotics; cells were harvested andpassaged when they reached 80% confluence.

Immunohistochemistry. IHC experiments were done on 5-μm formalin-fixedparaffin-embedded sections of normal pancreas (four cases) and pancreaticductal adenocarcinoma samples (12 cases; obtained from UCSF Cancer CenterTissue Core and from LifeSpan Biosciences, Inc.). The tissue sections weremounted on slides, treated with xylene and rehydrated using alcohol–watermixtures before heat treatment in 10 mM citric acid (pH 6.0) for antigen re-trieval. The endogenous peroxidase activity was quenched using 0.3% H2O2

solution. Following blocking with 10% serum, sections were incubated atþ4 °C overnight with primary human LRH-1 (hLRH-1) antibodies [LS-A2447(LifeSpan Biosciences) or HPA005455 (Sigma-Aldrich), 1∶200 dilution], thenrinsed and incubated with biotinylated anti-rabbit IgG (Vector Laboratories,1∶200 dilution) for 30 min at room temperature. After rinsing, the sampleswere treated with streptavidin and biotinylated horseradish peroxidase(Vectastain Elite kit, Vector Labs), stained with ImmPACT diaminobenzidine(Vector Labs) and counterstained with hematoxylin. Images were analyzedusing confocal microscopy (Nikon Eclipse Ti, magnification ×10 or ×40).The specificity of staining was confirmed using positive (liver) and negative(heart muscle tissue) controls.

siRNA Blocking. Pancreatic cancer cells AsPC-1 and COLO 357 were transfectedwith 10 nM of either of the two anti-LRH-1 siRNAs [corresponding to nucleo-

Fig. 5. Possible mechanisms of involvement of nuclear receptor LRH-1 in de-velopment of pancreatic cancer. Indicated are major transcriptional targetsof LRH-1 and regulatory pathways mediated by the receptor. The expressionof LRH-1 (yellow symbol) is controlled by the pancreas-specific master tran-scription factor PDX-1 (indicated in green) and via hedgehog signaling path-way that are both reactivated in pancreatic cancer. Activating mutations inthe LRH-1 gene and its regulatory regions might contribute to the receptorup-regulation. The activated receptor functions synergistically with β-catenin(shown in purple), activating multiple target genes regulating cell growthand proliferation (Cyc D1, Cyc E1, and C-Myc, indicated in gray) as well ascell differentiation (Oct4, Nanog, indicated in gray). The cumulative effectof the aberrant, LRH-1 mediated activation of these multiple targets in adultpancreas results in tumorigenesis and metastasis.

16930 ∣ www.pnas.org/cgi/doi/10.1073/pnas.1112047108 Benod et al.

Dow

nloa

ded

by g

uest

on

May

1, 2

021

tides 411–431 (siRNA-1) and 1504–1524 (siRNA-2) of hLRH-1] or control siRNAusing HiPerfect reagent (Qiagen) according to manufacturer’s protocol.Either siRNA targeting GFP or ALLStars siRNA labeled with Alexa Fluor 488(Qiagen) were used for controls, depending on the specifics of the experi-ments. The transfection efficiencies (75–90%) were evaluated by flow cyto-metry (BD LSR II flow cytometer, Becton Dickinson) and confirmed usingAllStars Hs Cell Death Control siRNA (Qiagen).

RNA Purification, cDNA Synthesis, and qPCR Analysis. For each sample, totalRNA was isolated using TRIzol (Invitrogen) according to manufacturer’sprotocol. Complementary DNA was synthesized from 1.5 μg of total RNA,at 42 °C for 60 min, using the SuperScript-II First-Strand Synthesis Systemfor RT-PCR (Invitrogen). Quantitative PCR amplification of mRNA for LRH-1and other targets was performed in triplicates using the Mx3005P Real-TimePCR System and the SYBR Green I dye (Stratagene) for detection. No Tem-plate Control and No Reverse Transcription Control were included in eachrun. The amplification curves were analyzed with the Mx3005P softwareusing the comparative cycle threshold method. All target mRNAs levels werenormalized relative to that of RS9 (internal control).

Western Blot Analysis. Cell lysate samples were separated by SDS PAGE andtransferred to nitrocellulose membranes. Western blot analyses were per-formed using goat anti-LRH-1 antibody and donkey antigoat horseradishperoxidase conjugated IGs (sc-5995 and sc-2033, Santa Cruz Biotechnology).The membranes were developed using enhanced chemiluminescence accord-ing to a standard protocol. The expression levels of LRH-1 were normalized to

that of β-actin (detected using goat anti-β-actin antibody; sc-1616, Santa CruzBiotechnology).

Cell Growth and Proliferation. Cultured cells were pelleted by centrifugationat 900 × g and resuspended in phenol red free medium supplemented with10% charcoal-stripped FBS (Hyclone) and 2% L-glutamine. Cells were platedin replicates of four (104 cells∕mL) in 96-well microtiter plates. At 24 h afterseeding, cells were transfected with either anti-LRH-1 or control siRNAs(10 nM) using HiPerfect reagent (Qiagen) according to manufacturer’sprotocol. At 96 h after the transfections, cell proliferation in each wellwas quantified using CellTiter-Glo assay (Promega).

Cell Cycle Analysis. Untreated cells AsPc-1 and cells treated with anti-LRH-1or control siRNAs were fixed with ethanol, stained with propidium iodide(Sigma), and analyzed using BD LSR II flow cytometer [Becton Dickinson;488-nm excitation, 562–588 nm emission (R-phycoerythrin (PE) fluorescence);single cells have been gated using PE-Area/PE-Width plot]. The cell cycle pro-files were analyzed with FlowJo V.8.8.6 (Flow Cytometry Analysis Software),using Dean–Jett–Fox algorithm.

ACKNOWLEDGMENTS. The authors thank Drs. S. Gysin and K. Shumate (HelenDiller Family Comprehensive Cancer Center, UCSF) for reagents and experi-mental suggestions, Drs. M. McMahon, M. Tempero (Cancer Center, UCSF),and M. Hebrok (Diabetes Center, UCSF) for helpful discussions, and R.Uthayaruban for technical assistance. This work was supported by NationalInstitute of Health Grants R01 DK078075 and R21 CA140751 (to R.J.F.).

1. Thayer SP, et al. (2003) Hedgehog is an early and late mediator of pancreatic cancertumorigenesis. Nature 425:851–856.

2. Pasca di Magliano M, et al. (2007) Common activation of canonical Wnt signaling inpancreatic adenocarcinoma. PLoS One 2:e1155.

3. Fayard E, Auwerx J, Schoonjans K (2004) LRH-1: An orphan nuclear receptor involvedin development, metabolism and steroidogenesis. Trends Cell Biol 14:250–260.

4. Fayard E, Schoonjans K, Annicotte J-S, Auwerx J (2003) Liver receptor homolog 1controls the expression of carboxyl ester lipase. J Biol Chem 278:35725–35731.

5. Kim JW, Peng N, Rainey WE, Carr BR, Attia GR (2004) Liver receptor homolog-1 reg-ulates the expression of steroidogenic acute regulatory protein in human granulosacells. J Clin Endocrinol Metab 89:3042–3047.

6. Clyne CD, et al. (2004) Regulation of aromatase expression by the nuclear receptorLRH-1 in adipose tissue. Mol Cell Endocrinol 215:39–44.

7. Gu P, et al. (2005) Orphan nuclear receptor LRH-1 is required to maintain Oct4 expres-sion at the epiblast stage of embryonic development. Mol Cell Biol 25:3492–3505.

8. Wagner RT, Xu X, Yi F, Merrill BJ, Cooney AJ (2010) CanonicalWnt/β-catenin regulationof liver receptor homolog-1 mediates pluripotency gene expression. Stem Cells28:1794–1804.

9. Heng JC, et al. (2010) The nuclear receptor Nr5a2 can replace Oct4 in the reprogram-ming of murine somatic cells to pluripotent cells. Cell Stem Cell 6:167–174.

10. Krylova IN, et al. (2005) Structural analyses reveal phosphatidyl inositols as ligandsfor the NR5 orphan receptors SF-1 and LRH-1. Cell 120:343–355.

11. Ortlund EA, et al. (2005) Modulation of human nuclear receptor LRH-1 activity byphospholipids and SHP. Nat Struct Mol Biol 12:357–363.

12. Whitby RJ, et al. (2011) Small molecule agonists of the orphan nuclear receptorssteroidogenic factor-1 (SF-1, NR5A1) and liver receptor homologue-1 (LRH-1,NR5A2). J Med Chem 54:2266–2281.

13. Lee JM, et al. (2011) A nuclear receptor dependent phosphatidylcholine pathway withantidiabetic effects. Nature 474:506–510.

14. Lee MB, et al. (2005) The DEAD-box protein DP103 (Ddx20 or Gemin-3) repressesorphan nuclear receptor activity via SUMO modification. Mol Cell Biol 25:1879–1890.

15. Lee YK, Choi YH, Chua S, Park YJ, Moore DD (2006) Phosphorylation of the hingedomain of the nuclear hormone receptor LRH-1 stimulates transactivation. J Biol Chem281:7850–7855.

16. Qin J, et al. (2004) Prospero-related homeobox (Prox1) is a corepressor of human liverreceptor homolog-1 and supresses the transcription of the cholesterol 7-a-hydroxylasegene. Mol Endocrinol 18:2424–2439.

17. Sablin EP, et al. (2008) The structure of corepressor Dax-1 bound to its target nuclearreceptor LRH-1. Proc Natl Acad Sci USA 105:18390–18395.

18. Botrugno OA, et al. (2004) Synergy between LRH-1 and beta-catenin induces G1cyclin-mediated cell proliferation. Mol Cell 15:499–509.

19. Sheela SG, Lee WC, Lin WW, Chung BC (2005) Zebrafish ftz-f1a (nuclear receptor 5a2)functions in skeletal muscle organization. Dev Biol 286:377–390.

20. Annicotte JS, et al. (2003) Pancreatic-duodenal homeobox 1 regulates expression ofliver receptor homolog 1 during pancreas development. Mol Cell Biol 23:6713–6724.

21. Quint K, et al. (2009) The expression pattern of PDX-1, SHH, patched and Gli-1 isassociated with pathological and clinical features in human pancreatic cancer.Pancreatology 9:116–126.

22. Heiser PW, et al. (2008) Stabilization of beta-catenin induces pancreas tumor forma-tion. Gastroenterololy 135:1288–1300.

23. Peterson G, et al. (2010) A genome-wide association study identified pancreaticcancer susceptibility loci on chromosomes 13q22 1, 1q32.1 and 5p15.33. Nat Genet42:224–228.

24. Schoonjans K, et al. (2005) Liver receptor homolog 1 contributes to intestinal tumorformation through effects on cell cycle and inflammation. Proc Natl Acad Sci USA102:2058–2062.

25. Annicotte JS, et al. (2005) The nuclear receptor liver receptor homolog-1 is an estrogenreceptor target gene. Oncogene 24:8167–8175.

26. Menssen A, Hermeking H (2002) Characterization of the c-MYC-regulated transcrip-tome by SAGE: Identification and analysis of c-MYC target genes. Proc Natl AcadSci USA 99:6274–6279.

27. Caldon CE, Musgrove EA (2010) Distinct and redundant functions of cyclin E1 and E2in development and cancer. Cell Div 5:2.

28. Stacy DW (2003) Cyclin D1 serves as a cell cycle regulatory switch in actively prolifer-ating cells. Curr Opin Cell Biol 15:158–163.

29. Ekholm SV, Reed SI (2000) Regulation of cyclin-dependent kinases in the mammaliancell cycle. Curr Opin Cell Biol 12:676–684.

30. Gysin S, Rickert P, Kastury K, McMahon M (2005) Analysis of genomic DNA alterationsand mRNA expression patterns in a panel of human pancreatic cancer cell lines. GenesChromosomes Cancer 44:37–51.

31. Jones S, et al. (2008) Core signaling pathways in human pancreatic cancers revealed byglobal genomic analyses. Science 321:1801–1806.

32. Harada T, et al. (2008) Genome-wide DNA copy number analysis in pancreatic cancerusing high-density single nucleotide polymorphism arrays. Oncogene 27:1951–1960.

33. Vadlamudi RK, Balasenthil S, Broaddus RR, Gustafsson JA, Kumar R (2004) Deregula-tion of estrogen receptor coactivator proline-, glutamic acid-, and leucine-rich protein-1/modulator of nongenomic activity of estrogen receptor in human endometrialtumors. J Clin Endocrinol Metab 89:6130–6138.

34. Chakravarty D, et al. (2010) Extranuclear functions of ER impact invasivemigration andmetastasis by breast cancer cells. Cancer Res 70:4092–4101.

35. Yamagata K, et al. (2009) Maturation of microRNA is hormonally regulated by anuclear receptor. Mol Cell 36:340–347.

36. Kumar MS, Lu J, Mercer KL, Golub TR, Jacks T (2007) Impaired microRNA processingenhances cellular transfomation and tumorigenesis. Nat Genet 39:673–677.

37. Li C, et al. (2007) Identification of pancreatic cancer stem cells. Cancer Res67:1030–1037.

38. Gidekel Friedlander SY, et al. (2009) Context-dependent transformation of adultpancreatic cells by oncogenic K-Ras. Cancer Cell 16:379–389.

39. Morris JP, Cano DA, Sekine S, Wang SC, Hebrok M (2010) Beta-catenin blocksKras-dependent reprogramming of acini into pancreatic cancer precursor lesions inmice. J Clin Invest 120:508–520.

40. Morris J, Hebrok M (2009) It’s free for all—insulin-positive cells join the group ofpotential progenitors for pancreatic ductal adenocarcinoma. Cancer Cell 16:359–361.

41. Coussens LM, Werb Z (2002) Inflammation and cancer. Nature 420:860–867.42. McColl SR, Paquin R, Menard C, Beaulieu AD (1992) Human neutrophils produced high

levels of the interleukin 1 receptor antagonist in response to granulocyte/macrophagecolony-stimulating factor and tumor necrosis factor alpha. J Exp Med 176:593–598.

43. Venteclef N, Delerive P (2007) Interleukin-1 receptor antagonist induction as anadditional mechanism for liver receptor hololog-1 to negatively regulate the hepaticacute phase response. J Biol Chem 282:4393–4399.

44. Mustea A, et al. (2008) Decreased IL-1 RA concentration in ascites is associated with asignificant improvement in overall survival in ovarian cancer. Cytokine 42:77–84.

45. Mueller M, et al. (2006) The nuclear receptor LRH-1 critically regulates extra-adrenalglucocorticoid synthesis in the intestine. J Exp Med 203:2057–2062.

46. Cox G (1995) Glucocorticoid treatment inhibits apoptosis in human neutrophils.Separation of survival and activation outcomes. J Immunol 154:4719–4725.

Benod et al. PNAS ∣ October 11, 2011 ∣ vol. 108 ∣ no. 41 ∣ 16931

APP

LIED

BIOLO

GICAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

May

1, 2

021